The ranking approach used by the Tampa Bay Times in their “America’s Worst Charities” articles is focusing on the high fees paid to telemarketers rather than just the overhead ratio. One of the threads of the online discussions is that their methodology is just focused on overhead ratios.
That is not the case. The overhead ratio conversation is a separate issue. The Tampa Bay Times approach focuses on the high cost of telemarketing as a fundraising technique and the issues related to that industry.
The reporters outlined their approach in a separate article here.
As I looked at bits and pieces of the data, I noticed several of the charities had respectable program ratios, so I decided to poke at the data.
A moment of background. The “overhead ratio” is typically calculated by adding the general & administrative costs to the fundraising costs on the 990. Then divide that sum by the total expenses. The complement of this number is the program ratio, which is calculated by dividing the program expenses on the 990 by total expenses.
To test my casual observation, I pulled key data from the 990s for the “top” 10 charities on the Tampa Bay list. The 990s are available at their website.
Here’s what I calculated for program ratios (program expense divided by total expense):
|1||Kids wish network||55%|
|2||Cancer fund of america||77%|
|3||Childrens wish foundation||52%|
|4||American Breast Cancer Found.||38%|
|5||Firefighters Charitable Found.||8%|
|6||Breast Cancer Relief Found.||57%|
|7||Int’l Union of Police Ass’n||18%|
|8||Nat’l Veterans Service Fund||29%|
|9||American Assn’ of State Troopers||23%|
|10||Children’s Cancer Fund of America||62%|
Some of those are really low.
One of those is actually reasonably high and several of them are still respectable.
So the ranking obviously has a very poor correlation to the overhead ratio.
There is another pattern in addition to what the reporters have identified.
I took the data for the top 10, which has a primary focus on the telemarketing costs, and looked at a couple more pieces of info.
Joint cost allocation – This is an area that is ripe for extended public discussion. I’ll put some numbers on the table now and explain the issue in a later post. I’ve been wanting to open up that discussion, but have lacked the time to develop what needs to be a long series of posts.
I pulled the costs allocated to program for joint costs and divided that by total expenses. This number represents the effort contained in cold calls that furthers the educational or programmatic efforts of the charity. Again, that gives the percent of total expenses allocated to program for the ‘educational’ or ‘program’ component of what would otherwise be fundraising costs.
Gifts in kind – This is an area that has already received extensive public discussion.
To figure out the impact on the top 10 from donated items, I pulled the noncash gifts from the revenue page of the 990 and divided that by total expenses. This is a rough calculation because it ought to be adjusted for changes in inventory, but is sufficient for the moment.
Then I put those numbers together to see what the picture looks like.
Let me describe the following table in words:
- Start with the total program ratio (prog cost / total exp).
- Subtract the joint costs that are allocated to program, expressed as a percent of total expenses.
- Subtract the GIK donations, expressed as a percent of total expenses.
- This leaves the portion of total expenses in the program category that are neither allocated joint costs (i.e. telemarketing) nor donated goods. Lacking a more technical description, the remaining number is the cash spent on program services as a percent of total expenses.
Now you can see an overall pattern.
I’ve not given any conclusions. Figure those out for yourself.
I’ve explained my methodology so you can revise it to your taste, ignore it, or explain why it is invalid. You can interpret the methodology on your own.
Your thoughts? Is my methodology invalid? Did I goof up on some of the calculations?
Comments are very welcome.